JBacteriology

Microbiol Immunol Infect fasciitis

of necrotizing
2005;38:430-435

Microbiology and factors affecting mortality

in necrotizing fasciitis
Yuag-Meng Liu1, Chih-Yu Chi2, Mao-Wang Ho2, Chin-Ming Chen2, Wei-Chih Liao2, Cheng-Mao Ho2,
Po-Chang Lin2, Jen-Hsein Wang2

1
Department of Internal Medicine, St. Joseph Hospital, Yulin; and 2Section of Infectious Diseases,
Department of Internal Medicine, China Medical University Hospital, Taichung, Taiwan

Received: May 4, 2005 Revised: July 12, 2005 Accepted: August 19, 2005

Necrotizing fasciitis is a life-threatening soft-tissue infection primarily involving the superficial fascia. This study
investigated the microbiologic characteristics and determinants of mortality of this disease. The medical records
of 87 consecutive patients with a diagnosis of necrotizing fasciitis from 1999 to 2004 were retrospectively reviewed.
A single pathogen was identified as the infectious agent in 59 patients (67.8%), multiple pathogens were identified
in 17 patients (19.6%), and no organism was identified in 11 patients (12.6%). Klebsiella pneumoniae, identified in
17 patients, was the most commonly isolated species. The most common comorbidity was diabetes mellitus (41
patients; 53.2%). Multivariate logistic regression analysis showed that more than 1 comorbidity, thrombocytopenia,
anemia, more than 24 h delay from onset of symptoms to surgery and age greater than 60 were independently
associated with mortality. This study found that K. pneumoniae was the most common cause of necrotizing fasciitis.
Early operative debridement was independently associated with lower mortality.

Key words: Debridement, Klebsiella pneumoniae, necrotizing fasciitis

Necrotizing fasciitis is a life-threatening soft tissue
infection primarily involving the superficial fascia [1-
9]. It is perhaps the most aggressive form of soft tissue
infection [2-7]. Meleney reported the first major series
of patients with necrotizing fasciitis in 1924, which was
a group of 20 cases observed in China [1]. His study
disclosed beta-hemolytic streptococci infection in all
of the patients, and he named the disease acute
streptococcal gangrene.
The term necrotizing fasciitis was first used by
Wilson in 1952 [2]. The consensus was that necrotizing
fasciitis was an infection involving the superficial
fascia and the subcutaneous tissue with very minor
superficial epidermal or mucosal focus as a portal of
bacterial entry. The progression of the disease is
fulminant [2-4]. The prognosis hinges on accurate
diagnosis and immediate institution of appropriate
treatment [4-7]. We had observed that a high portion of
cases of necrotizing fasciitis in our institute were caused
by Enterobacteriaceae but that epidemiologic data were
lacking. This study analyzed the microbiologic spectrum
Corresponding author: Jen-Hsein Wang, Head of Infectious Diseases,
Department of Internal Medicine, China Medical University
Hospital, No. 2 Yude Rd., North District, Taichung City 404, Taiwan.
E-mail: jenhsien@www.cmuh.org.tw
of this condition at our hospital. The factors affecting
mortality associated with necrotizing fasciitis were also
evaluated.
Materials and Methods
A computer-generated search in the database of the
Medical Records Department was used to identify
patients with operative findings suggesting necrotizing
fasciitis or computed tomography (CT) revealing
asymmetric fascia thickening [3] who were treated
between 1999 and 2004.
Diagnosis of necrotizing fasciitis was made based
on the following characteristics at operative exploration:
1) the presence of grayish necrotic fascia; and 2) easy
separation of the superficial fascia from the underlying
tissues during surgery. Permanent histopathologic tissue
examination was used to confirm the diagnosis [4].
Data collected from records of each patient included
age, gender, location of infection, number and type
of comorbid illnesses, portal of entry of infection,
symptoms at admission (including the time to access
medical care), vital signs, and physical, radiographic,
and laboratory findings at the time of admission. The
results of microbiologic cultures of tissue samples

430
 Liu et al

obtained at the first operative debridement were Table 1. Demographic, clinical, biologic, and radiographic data
analyzed. Data on delay from initial presentation of of patients with necrotizing fasciitis
symptoms to operative treatment, initial antibiotics Variable n or mean (range [%])
used, number of operative debridements, performance Gender
of amputation, duration of hospitalization, and the in- Male 65 (74.7)
hospital mortality rate were also recorded. The anatomic Female 22 (25.3)
site of infection was classified as either central (trunk, Age (years) 53.4 (1-86)
back, or groin) or peripheral (upper and lower limbs). Comorbidity
Statistical analyses were performed using Statistical Diabetes mellitus 44 (50.6)
Liver cirrhosis 25 (28.7)
Package for the Social Sciences software. Comparisons
Alcoholism 18 (20.7)
of proportions were made using Pearsons chi-squared
Immunosuppressant agent 6 (6.9)
test to identify univariate differences among defined Chronic renal insufficiency 2 (2.6)
variables with respect to mortality. Students t test Cancer 2 (2.3)
was used for the analysis of continuous variables. The Human immunodeficiency virus 1 (1.3)
significance of factors with a possible influence on Aplastic anemia 1 (1.3)
mortality was evaluated using a logistic regression No comorbidities 16 (18.3)
approach by means of a backward stepwise selection Location
Head and neck 5 (5.7)
procedure. A p value of 0.2 was chosen as the criterion
Trunk 10 (11.5)
to judge the entry and removal of variables at each step Inguinal area 9 (10.3)
of the regression procedure. Then, the final model was Upper limb 6 (6.9)
constructed to determine the factors independently Lower limb 57 (65.5)
associated with mortality. Findings on admission
Temperature >38C 21 (24.1)
Results Hypotension 28 (32.2)
Amputation performed 9 (10.3)
Gas on radiograph 4 (4.6)
The medical charts of 113 patients identified as
Mortality 30 (34.5)
having a diagnosis of necrotizing fasciitis at our institute Total WBC count (k/mL) 13.5 (0.7-50.6)
during the study period were obtained for review. C-reactive protein (mg/dL) 15.0 (0.9-60.8)
Twenty six of these patients were excluded because their Number of debridements 3.0 (0-23)
histopathologic tissue examinations did not favor the Duration of hospitalization (days) 33.7 (1-163)
diagnosis. The remaining 87 patients were included in Abbreviation: WBC = white blood cell
this study. Table 1 summarizes the demographic
characteristics and the clinical, biochemical and radio- Cultures of tissue specimens obtained during the
graphic findings in these patients at admission. The first operative debridement were analyzed. A single
major presenting symptoms included swelling (69 pathogen was isolated in 59 patients (67.8%), more
patients, 89.6%), pain (68 patients, 88.3%), erythema than 1 pathogen was identified in 17 patients (19.6%),
(49 patients, 63.6%), local heat (22 patients, 36.3%), and no organism was identified in 11 patients (12.6%).
and bullae (22, 26.0%). Other findings included skin K. pneumoniae, identified in 17 patients, was the most
induration, crepitus, fluctuance, and skin necrosis as well common pathogen isolated.
as sensory and motor deficits. The bacteriology of monomicrobial necrotizing
The portal of entry of infection was identified in fasciitis is summarized in Table 2. Among the 59
50 patients. The source of necrotizing fasciitis could not patients with monomicrobial necrotizing infections,
be found in 39 patients. Trauma was the most common Gram-positive bacteria were isolated in 25 patients.
portal of entry, which was found in 39 patients (44.8%). Staphylococcus aureus was the most common Gram-
Nine patients had necrotizing fasciitis spread from a local positive isolate, found in 13 patients. Gram-negative
infection. These infections included peritonsillar abscess bacteria were isolated in 34 patients. Nineteen of these
(3 patients), angioedema of the lip (1 patient), empyema isolates were members of Enterobacteriaceae. K.
thoracis (1 patient), diverticulitis (1 patient), site of pre- pneumoniae was the most common Gram-negative
existing bed sore (1 patient), and Klebsiella pneumoniae pathogen isolated. Aeromonas spp. were isolated in
metastatic infection (2 patients). 7 patients, all of whom died. Vibrio vulnificus was

431
Bacteriology of necrotizing fasciitis

Table 2. Bacteriology of monomicrobial necrotizing fasciitis disease. Mortality was significantly increased in patients
No. with 2 or more comorbidities, such as diabetes mellitus
combined with liver cirrhosis.
Gram-positive
Staphylococcus aureus 13
Table 4 compares the laboratory data between
Group A Streptococcus 8 survivors and non-survivors. Conditions significantly
Group B Streptococcus 2 associated with mortality included thrombocytopenia,
Group F Streptococcus 1 abnormal liver function, and low serum albumin level.
Viridans group streptococci 1 Patients who received emergent debridements at least
Gram-negative 24 h before the onset of symptoms had a lower mortality
Klebsiella pneumoniae 13
rate than those who had delayed operations (26.0%
Escherichia coli 2
versus 45.9%, p=0.069). Although mortality was lower
Enterobacter cloacae 1
Citrobacter 1
when the isolated organisms were sensitive to initial
Proteus spp. 2 antibiotics treatment, this difference was not significant
Aeromonas 7 (31.9% versus 37.5%, p=0.654).
Vibrio vulnificus 3 Because of the large number of potentially inter-
Acinetobacter baumannii 2 dependent parameters examined in this retrospective
Pseudomonas aeruginosa 3 study, logistic regression analysis was performed to
Mixed infection 17
assess the independent effect of variables on mortality.
Variables that were significant in the univariate analysis
isolated in 3 patients, 2 of whom died. Three patients (p<0.2) were selected for inclusion in the first step of the
had Pseudomonas aeruginosa necrotizing fasciitis and stepwise regression model. These parameters included age,
all of them survived. shock, fever, cirrhosis, 2 or more underlying conditions,
Mixed infection was documented in 17 patients. use of immunosuppressant, low hemoglobin, decreased
Enterobacteriaceae were the most common bacteria platelet count, acute renal failure, decreased serum albumin
isolated in polymicrobial necrotizing fasciitis, as in level, delay of more than 24 h from symptom onset to
monomicrobial infections. Among the Enterobacteriaceae surgery, number of times of debridement and bacteremia.
isolates, Escherichia coli and K. pneumoniae were The multivariate logistic regression analysis revealed
the most common species, followed by Enterobacter that risk of death was independently associated with more
cloacae, Morganella morganii and Proteus vulgaris. P. than 1 underlying condition, thrombocytopenia, anemia,
aeruginosa was isolated in 3 patients, and Acinetobacter delay of more than 24 h from symptom onset to surgery,
spp. in 2 patients. Among anaerobic organisms isolated, and age greater than 60 years.
Bacteroides fragilis was the most common. Prevotella
spp. were isolated in 2 patients, Veillonella in 1 patient, Discussion
Propionibacterium propionicum in 1 patient, and
Eikenella corrodens in 1 patient. Giuliano et al were the first to divide bacteriologic
Amputation was performed to control the infection culture results in necrotizing fasciitis into 2 distinct
in 9 patients (11.7%). Patients underwent a mean of groups in a study of 16 cases reported in 1977 [5].
3.0 debridements (range, 0 to 15 debridements) to From their point of view, aerobic organisms other
control the infective process. The average duration of than Group A Streptococcus alone could not cause
hospitalization was 33.7 days (range, 1 to 163 days). necrotizing fasciitis. Subsequent studies revealed
Thirty patients (32.5%) died. Table 3 shows that mixtures of aerobes and anaerobes could act
the relationship of pre-existing characteristics and synergistically, thus affecting the virulence of
medical condition to mortality. Death was significantly necrotizing fasciitis [5-10].
associated with age greater than 60 years. Mortality was It was also demonstrated that solitary aerobic
not significantly influenced by other physical findings organisms other than Group A Streptococcus could cause
during hospitalization such as pain, swelling, erythema necrotizing fasciitis. Howard et al reported 18 patients
or local heat. with necrotizing fasciitis caused by Vibrio spp. in 1985
Although diabetes mellitus was the most common [11]. These infections usually occur in apparently
pre-existing medical condition, it was not significantly insignificant wounds exposed to sea water or fish
associated with mortality unless it occurred with other [10-14]. The mortality rate in their study was 33.3%.

432
 Liu et al

Table 3. Mortality in groups defined by pre-existing characteristics and medical conditions

Survivors Non-survivors Mortality rate (%) p

Initial presentation
Age (years)
<60 36 14 28.0
60 21 16 43.2 0.173
Gender
Female 15 6 28.6
Male 42 24 36.4 0.802
Shock
No 42 17 28.8
Yes 15 13 46.4 0.147
Disturbed consciousness
No 51 24 32.0
Yes 6 6 50.0 0.326
Fever
No 40 26 39.4
Yes 17 4 19.0 0.116
Bullae
No 44 19 30.2
Yes 13 11 45.8 0.210
Trauma
No 25 16 39.0
Yes 32 14 30.4 0.499
Pain
No 6 4 40.0
Yes 51 26 33.8 0.732
Swelling
No 4 5 55.6
Yes 53 25 32.1 0.265
Local heat
No 38 18 32.1
Yes 19 12 38.7 0.639
Erythema
No 18 13 41.9
Yes 39 17 30.4 0.348
Peripheral
No 15 8 34.8
Yes 42 22 34.4 0.972
Cormorbidity
Diabetes mellitus
No 31 12 27.9
Yes 26 18 40.9 0.261
Cirrhosis
No 44 18 29.0
Yes 13 12 48.0 0.134
Alcoholism
No 44 25 36.2
Yes 13 5 27.8 0.589
Immunosuppressant
No 55 26 32.1
Yes 2 4 66.7 0.176
2 comorbidities
No 39 9 18.8
Yes 18 21 53.8 0.001

433
Bacteriology of necrotizing fasciitis
Table 4. Relationship of laboratory data with mortality
Variable (mean SD) Survivors
Hemoglobin (mg/dL) 11.2 2.6
WBC count (/mL) 13,442 8790
Platelet (k/mL) 248 162
Bilirubin (mg/dL) 3.1 2.7
AST (U/L) 110 277
ALT (U/L) 64 171
BUN (mg/dL) 24 17
Creatinine (mg/dL) 1.3 0.7
Albumin (mg/dL) 2.1 0.9
CRP (mg/dL) 14 14.9
Abbreviations: SD = standard deviation; WBC = white blood cell; AST = aspartate aminotransferase; ALT = alanine aminotransferase;
BUN = blood urea nitrogen; CRP = C-reactive protein
In this series, Vibrio spp. were isolated in 3 patients.
Two of these patients had cirrhosis and died despite
debridements and adequate antibiotic therapy.
Another highly virulent bacterium which can cause
necrotizing fasciitis is Aeromonas spp. This organism
was the etiologic agent in 7 patients in our study and all
of them died. Aeromonas hydrophila has been described
as the cause of the necrotizing fasciitis in patients with
immunosuppression, burns or trauma in an aquatic
setting [15]. Patients with liver cirrhosis or malignancy
had a higher mortality rate within 7 days after admission
[16]. Among the 7 patients with necrotizing fasciitis
caused by Aeromonas spp. in this series, 4 had cirrhosis,
3 had diabetes mellitus, and only 1 patient had no
underlying disease. The prevalence of these underlying
conditions is consistent with previous studies [15-18].
Highly virulent bacteria such as many of the
facultative bacteria and P. aeruginosa, just like the group
A streptococci, can cause necrotizing fasciitis alone,
especially in high-risk patients [19].
Enterobacteriaceae were the most common group
of bacterial organisms isolated in this series. A single
Enterobacteriaceae was isolated in 19 patients. Most
patients with necrotizing fasciitis caused by solitary
Enterobacteriaceae had underlying disease including
diabetes mellitus in 13, cirrhosis in 4, both diabetes and
cirrhosis in 3, while only 5 patients had no underlying
disease. K. pneumoniae, found in 17 patients, was the
most common pathogen isolated in this series, unlike a
previous series in which beta-hemolytic Streptococcus
was the most common single pathogen isolated [12].
The association of K. pneumoniae necrotizing fasciitis
with liver abscess and endogenous endophthalmitis of
the eye has recently been highlighted [20-22]. Two of
the patients with K. pneumoniae necrotizing fasciitis in
this series had metastatic infection. Necrotizing fasciitis
434
Non-survivors p
10.3 2.7 0.177
13,745 9730 0.883
129 123 0.01
5.8 5.4 0.68
88 58 0.72
46 24 0.637
48 41 <0.001
2.9 2.9 <0.001
1.6 0.5 0.05
15.9 13.1 0.72
was not the initial presentation in these 2 patients,
both of whom had K. pneumoniae bacteremia and
diabetes mellitus. One of the patients also had acute
pyelonephritis and the other had liver abscess.
Necrotizing fasciitis developed later in the course of
hospitalization. These cases emphasize the need for
including necrotizing fasciitis in the differential
diagnosis of patients with K. pneumoniae infection who
present with limb swelling.
Necrotizing fasciitis was caused by Group B
Streptococcus in 2 patients in this series, both of whom
had diabetes mellitus. The reported predisposing factors
to this etiology include obstetric complications in
postpartum adult females and in infants as well as
diabetes [23-27].
The somewhat high rate of monomicrobial isolation
in this study might have been due to the use of relatively
unsophisticated techniques for collection, transfer or
culture of anaerobic specimens. The empirical use of
antibiotics with activity against anaerobes antibiotics,
however, had no impact on mortality rate. This may
imply that anaerobes were easily eliminated by
aggressive debridements. Anti-anaerobic antibiotics
were used for more than 6 h before obtaining culture in
only 12 of the 87 patients in this series. The impact of
empirical antibiotics on the results of culture was thus
likely to be limited.
Multivariate logistic regression analysis revealed
that more than 1 comorbidity, thrombocytopenia,
anemia, delay of more than 24 h from symptom onset
to surgery, and age greater than 60 years independently
affected survival. These findings are similar to the
study of Wong et al, who reported that advanced age, 2
or more associated comorbidities, and a delay in surgery
of more than 24 h adversely affected outcome [14].
The retrospective nature of this study, however, makes

these findings inconclusive. Age and underlying disease
reflect the immune status of patients. The presence
of chronic, debilitating disease may contribute to the
uncontrollable nature of both local and systemic
infection. Early debridement offers the only opportunity
to correct these problems.
The paucity of cutaneous findings early in the
course of necrotizing fasciitis made early diagnosis
difficult in this series. The use of broad-spectrum
empirical antibiotics to cover staphylococci, streptococci,
and Enterobacteriaceae from the beginning appears
indicated based on the bacteriologic findings. The need
for a high index of suspicion of necrotizing fasciitis,
and prompt early debridement in identified cases cannot
be overemphasized.
References
1. Meleney FL. Hemolytic streptococcus gangrene. Arch Surg
1924;9:317-64.
2. Wilson B. Necrotizing fasciitis. Am Surg 1952;18:416-31.
3. Wysoki MG, Santora TA, Shah RM, Friedman AC. Necrotizing
fasciitis: CT characteristics. Radiology 1997;203:859-63.
4. Majeski J, Majeski E. Necrotizing fasciitis: improved survival
with early recognition by tissue biopsy and aggressive surgical
treatment. South Med J 1997;90:1065-68.
5. Giuliano A, Lewis F Jr, Hadley K, Blaisdell FW. Bacteriology
of necrotizing fasciitis. Am J Surg 1977;134:52-7.
6. Pessa ME, Howard RJ. Necrotizing fasciitis. Surg Gynecol
Obstet 1985;161:357-61.
7. Rea WJ, Wyrick WJ Jr. Necrotizing fasciitis. Ann Surg 1970;
172:957-64.
8. Stone HH, Martin JD Jr. Synergic necrotizing cellulitis. Ann
Surg 1972;175:702-11.
9. Wang KC, Shih CH. Necrotizing fasciitis of the extremities.
J Trauma 1992;32:179-82.
10. Kingston D, Seal DV. Current hypotheses on synergistic
microbial gangrene. Br J Surg 1990;77:260-4.
11. Howard RJ, Pessa ME, Brennaman BH, Ramphal R.
Necrotizing soft-tissue infections caused by marine vibrios.
Surgery 1985;98:126-30.
12. Elliott D, Kufera JA, Myers RA. The microbiology of
necrotizing soft tissue infections. Am J Surg 2000;179:361-6.
13. Fontes RA Jr, Ogilvie CM, Miclau T. Necrotizing soft-tissue
infections. J Am Acad Orthop Surg 2000;8:151-8.
14. Wong CH, Chang HC, Pasupathy S, Khin LW, Tan JL,
Liu et al
Low CO. Necrotizing fasciitis: clinical presentation,
microbiology and determinants of mortality. J Bone Joint Surg
Am 2003;85-A:1454-60.
15. Angel MF, Zang F, Jones M, Henderson J, Chapman SW.
Necrotizing fasciitis of the upper extremity resulting from a
water moccasin bite. South Med J 2002;95:1090-4.
16. Lau SM, Peng MY, Chang FY. Outcomes of Aeromonas
bacteremia in patients with different types of underlying disease.
J Micorbiol Immunol Infect 2000;33:241-7.
17. Ko WC, Lee HC, Chuang YC, Liu CC, Wu JJ. Clinical features
and therapeutic implications of 104 episodes of monomicrobial
Aeromonas bacteremia. J Infect 2000;40:267-73.
18. Minnaganti VR, Patel PJ, Iancu D, Schoch PE, Cunha BA.
Necrotizing fasciitis caused by Aeromonas hydrophila. Heart
Lung 2000;29:306-8.
19. Ou LF, Yeh FL, Fang RH. Bacteriology of necrotizing fasciitis:
a review of 58 cases. Zhonghua Yi Xue Za Zhi (Taipei) 1993;
51:271-5.
20. Ho PL, Tang WM, Yuen KY. Klebsiella pneumoniae necrotizing
fasciitis associated with diabetes and liver cirrhosis. Clin Infect
Dis 2000;30:989-90.
21. Hu BS, Lau YJ, Shi ZY, Lin YH. Necrotizing fasciitis associated
with Klebsiella pneumoniae liver abscess. Clin Infect Dis 1999;
29:1360-1.
22. Dylewski JS, Dylewski I. Necrotizing fasciitis with Klebsiella
liver abscess. Clin Infect Dis 1998;27:1561-2.
23. Holmstrom B, Grimsley EW. Necrotizing fasciitis and toxic
shock-like syndrome caused by group B streptococcus. South
Med J 2000;93:1096-8.
24. Gardam MA, Low DE, Saginur R, Miller MA. Group B
streptococcal necrotizing fasciitis and streptococcal toxic
shock-like syndrome in adults. Arch Intern Med 1998;158:
1704-8.
25. Reyzelman AM, Armstrong DG, Vayser DJ, Hadi SA, Harkless
LB, Hussain SK. Emergence of non-group A streptococcal
necrotizing diabetic foot infections. J Am Podiatr Med Assoc
1998;88:305-7.
26. Riefler J 3rd, Molavi A, Schwartz D, DiNubile M. Necrotizing
fasciitis in adults due to group B streptococcus. Report of a
case and review of the literature. Arch Intern Med 1988;148:
727-9.
27. Tang WM, Ho PL, Yau WP, Wong JW, Yip DK. Report of 2
fatal cases of adult necrotizing fasciitis and toxic shock
syndrome caused by Streptococcus agalactiae. Clin Infect Dis
2000;31:E15-7.
435